The idea that multiplication of electrons by direct collisions with gas atoms can account for the starting of discharges has been found to be untenable in neon over a wide range of pressure and wave-lengths. Also, the dictum that wall processes can be neglected when the gas pressure is high appears to be a fallacy. The experiments leading to these conclusions were done with cylindrical glass vessels with plane ends and external electrodes and uniform alternating electric fields. With pressures from 2 to 200 mm Hg, as the wave-length was varied between 10 and 10
7
m, the starting field showed three plateaux, the lowest at short wave-lengths. Here for
p
> 50mm Hg the field per unit pressure was found to be constant and very low, namely, 0·6 V/cm mm Hg. Using even the most favourable energy distribution, the fraction of electrons exceeding ionization energy of about 21 eV is much too small to give electron multiplication by collision which can balance the losses by diffusion. However, there is a much larger fraction of electrons which can excite neon atoms to resonance or metastable levels. Thus a new picture emerges; a chance electron which is accelerated by the field excites neon atoms to about 16 eV. The emitted quanta which fall on the glass walls release photo-electrons which join the first electron, etc.; hence, not only are electrons multiplied but also quanta. When the concentration of excited atoms has become sufficiently great, the large number of slow electrons with energies > 5 eV can ionize the excited gas. Thus the starting field corresponds not to ionization by collision but to the onset of multiplication of quanta and photo-electrons during the first stage of the breakdown. The theory given leads to a relation between the starting field and its wave-length, the gas pressure, the size of the vessel, the nature of the gas and of the wall. Good numerical agreement with observations is found, the constants being taken from known atomic data. The concept of the electron multiplication sustained by quanta may have a bearing on other types of discharge in different gases.
A Method for the Measurement of Carbon Dioxide Desorption from Coal in the Elevated Pressure Range